 |
- Always remember the fundamental goal of writing a scientific paper:
Intimidation and not communication; if you explain your ideas
clearly, people will steal them whereas if they look complex they will
admire you.
- Do not let your creativity be hemmed by a rigid structure; a good
theory will explain itself. Always remember that a scientific exposition
is not a work of literature; it is a work of genius.
- Show your creativity by not following the whims and fashions of
previous notation. For example, it is perfectly acceptable to denote
inputs by y and outputs by x, sample size by
e and accuracy by m. As a side effect,
your complexity results might look a lot better.
- Choose your notation carefully. You can safely assume that all your
readers will know about strict typing and operator precedence. It is
perfectly acceptable that x denotes a number, a vector, a matrix, a
string, a list, a graph, an algebra, a topological space, a ring, and a
page number at the same time. Always remember that the Latin alphabet has
only got 26 letters so do not squander them.
- Your paper will have a more sophisticated look if you exclusively use
the Greek, Gothic and Hebrew alphabet.
- Do not feel obliged to have any connection between the introduction
and the main body; after all smart reviewers will only
read the introduction anyway.
- There are circumstances in which communication is the goal of
your paper. In such cases, use Shannon's information theory results which
show that redundancy can improve the communication accuracy. Do this, for
example, by simply repeating whole paragraphs of your paper. This is easy
with the cut-and-paste functionality of modern editors.
- The most important part of your paper are the references. Do not
squander a citation on someone whose favour you do not wish to curry.
Similarly, it is essential liberally cite those researchers who you
conjecture might be reviewers of your paper: this will be the
first thing the reviewers will
check.
- Make sure that you do not acknowledge one of your co-authors; it gives
away too much of the paper's history.
- Demonstrate your total concentration on the deep science within your
paper by outrageous violation of the rules of grammar; for example, omit
verbs from key sentences, deprecate the use of the definite article and
utterly ignore conventions of punctuations and capitalisation.
- If you managed to become a big-shot then simply assign the task of
writing to your student. Make sure he
gets acknowledged.
- Never write an open question in your paper that you have not already
answered. You will only get asked by the editor
to solve it.
- If you are at a loss as to what to write about, a good strategy is to
find an old algorithm in the literature; rename it and obfuscate it (by
change of notation - see above). If necessary introduce inefficiencies
which you will be able to remove by an "improved algorithm" in a later
paper. After a decent interval (a year will do) you can write another
paper
- Think about a good structure of your paper and try to use consistent
and well known notation. Always keep in mind that you write the paper in
order to communicate your idea to the readers. Adhere to the rules of
grammar and avoid typographic errors.
|
